6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
Does Your Occupation Make You Sick?
Some Things We Know from the National Longitudinal Survey of Youth
Mark Allyn
Department of Management & Information Systems
Montclair State University
Author Notes:
Correspondence concerning this manuscript should be sent to: Mark Allyn, Department of
Management & Information Systems, School of Business, Montclair State University,
Montclair, NJ 07043; Phone: (973) 655-7708; E-mail: allynm@montclair.edu;
Fax: (973) 655-7678.
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
1
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
Does Your Occupation Make You Sick?
Some Things We Know from the 1979 National Longitudinal Survey of Youth
Abstract
Numerous epidemiological studies have shown a negative and significant correlation between
Socioeconomic Status (SES) and worker health. Several theories have been advanced that offer explanations for
why this might be true. This paper investigates the effects of variables included in these theories on health
among participants of the 1979 National Longitudinal Survey of Youth. We present a structural equations
model including these variables and human capital. We show that the nature of the job, not only its pay, but
also other important features such as benefits and promotions, have relatively small effects on health compared
to those human capital variables such education, abilities, and personal sense of mastery. Depression is a strong
predictor of morbidity, and human capital variables modulate depression and obesity. One implication for
employers who seek to improve worker health is to spend more effort recruiting individuals who are high on
self-perceived competence since this correlates fairly well with education and abilities.
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
2
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
Does Your Occupation Make You Sick?
Some Things We Know from the 1979 National Longitudinal Survey of Youth
Social scientists and public health workers have focused increasing attention on the relationship
between socioeconomic status (SES) and morbidity and mortality. Socioeconomic status, a sociological
construct that involves education, income, and job prestige, has been shown repeatedly to be inversely related to
morbidity and mortality (Rodgers, 1979). This robust finding crops up in numerous studies across many
societies around the world (Lynch & Kaplan, 1997).
Somewhat surprisingly, business people have shown little interest in the phenomenon even though
worker pay is by far the largest contributor to individual and household income. The SES data on health
suggests that there is a consistent relationship between what people get paid and how healthy they are. Most of
the business world’s attention seems devoted to controlling benefits costs, satisfying U.S. OSHA requirements,
absenteeism, substance abuse, and the like, with occasional forays into “wellness programs” whose value is
somewhat indeterminate. What we as business people have paid scant attention to is the health outcomes we
can expect from our pay policies. For example, pay-for-performance is a much lauded productivity enhancer,
but is it good for the health of our workers? We simply do not know.
While the finding that SES is monotonically related to morbidity and mortality is very stable, the
theory connecting them is in its infancy. It is easy enough to understand that absolute poverty exposes one to
significant health hazards, but this doesn’t seem to accord well with the steady and continuous improvement in
health that is observed. Absolute poverty might also be related to physical injury and accidents on the job, but
surely the jobs of clerical workers are no more physically dangerous than those of their managers, but the
managers consistently enjoy better health (Marmot, et al. 1997). The explanation favored by many in the public
health field, is that people with lower SES are psychologically deprived of feelings of self-worth and dignity,
that this deprivation somehow is stressful, that the stress in turn provides an environment in which disease
vectors can flourish (Marmot, 2004; Lynch and Kaplan, 1997). Parenthetically we note that very little has been
done with the Education and Esteem components of SES.
A second explanation involving SES is (“Effort-reward imbalance”) that people who are paid relatively
less than others are likely to work in jobs where the pay is not commensurate with the level of effort required by
employers (Siegrist, 1996). This input/output imbalance is stressful and the sufferer thus is prone to disease
factors that are encouraged by stressed somatic conditions. The third explanation (“demand control”) also relies
on the idea of stress, but in this explanation the trigger is a lack of control over job activities while
simultaneously being driven hard by the business (Theorell and Karasek, 1996). Empirical work by Theorell, et
al. (1998) and Bosma, et al. (1998) has found support for both the latter explanations of the
SES/morbidity/mortality relationship.
Both the effort/Reward and Demand Control explanations implicate other characteristics of the workers
job as contributing to disease besides just earnings. Demand control gets at the extent to which the job allows
the worker to exercise discretion and also the extent to which that discretion is effective. Thus, giving workers
adequate job training and structuring job activities such that the worker can exercise choice are important in
health outcomes. Similarly, the Effort/Reward imbalance theory draws attention to workers intrinsic rewards
from jobs as well as extrinsic pay and social esteem. Rewards can come in the form of promotions, additional
vacation time, recognition, and so forth. Consequently, both of these models take one well beyond a simple
model of economic pay and its distribution and resulting health problems.
In this paper we highlight SES and also other variables that should reflect differences in Effort-Reward
Imbalance and to Demand Control:
A common theme in all three explanations is the idea that stress is the proximal cause of morbidity, and
that social conditions indexed by SES can induce harmful biological conditions via stress. This connection with
biology actually has been demonstrated in primates by Sapolsky (1990) and also by Shively and Clarkson (1988,
1994), and in humans by Adler and her colleagues (2000).
Of course, SES is not the only significant behavioral and social agent in the etiology of disease. Many
variables including obesity, exercise, alcohol, tobacco, and other drug use, as well as living conditions, are
implicated in morbidity. Even IQ appears to be a statistically significant correlate of health outcomes, although
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
3
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
as we will show in this paper, the connection appears indirect. A consistent finding, and one that the present
paper will highlight, is that psychological depression plays a prominent role. Many variables are intertwined
and untangling them seems to be a matter of high priority, especially since so many socially conscious medical
professionals quickly opt for some kind of income redistribution scheme to ameliorate the problem (Marmot,
2003; Lynch & Kaplan, 1997)
Despite nearly three decades of work in this field (see Feinstein, 1993), much of it done with great
meticulousness by competent statisticians and medical professionals, there does not seem to be a structural
model for morbidity that relates SES and other of the frequently mentioned agents to human morbidity. The
present paper has set itself the task of developing just such a model. To that end we apply structural equation
modeling methods to a large American longitudinal data set, the National Longitudinal Survey of Youth, begun
in 1979.
Data
The data we use are taken from the 1979 National Longitudinal Survey of Youth. The NLSY79 is
funded by the U.S. Bureau of Labor Statistics and is managed by the Center for Human Resource Research at
the Ohio State University. The cohort was formed from a probability sample 12,686 young people between the
ages of 14 and 22 in 1979, with an average age of 17.9 years. The respondents are re-interviewed annually or
biannually. As of 2002, the year in which our health inventory data were collected, there were 7724 available
respondents between the ages of 37 and 45 with an average of 40.9 years. Not all these respondents were
employed, however, and other sources of non-response reduced the sample available for statistical analysis to
about 3500 respondents with complete response profiles for some of the analysis. Since the sample had on
average just entered its forties, the incidence of serious morbidity had only begun to become evident.
Depending upon the specific morbidity measure, the incidence of morbidity ran between 14 and 17 percent of
the sample. Many health-affecting behaviors such as obesity, lack of physical exercise, alcohol use, smoking,
and non-prescription drugs were present, as we shall see. One of the most important data elements coded into
the NLSY79 is the U.S. 3-digit occupation code. Based on this code (for 1980) we were able to assign job
prestige ratings to the respondents.
Male respondents comprised about 49 percent of the sample. The sample self-identified as about 70
percent white.
Variables
The following variables were either extracted directly from the survey raw data or else created,
typically by factor analysis. We have classified them into groups. The first group is characteristics of the
worker that he/she brings to the workplace. They represent human capital, and include abilities-related
cognitive capacities such as AFQT and social assets such as marriage and family. The second category are
variables that are job characteristics, including pay, prestige, job hours flexibility, and a number of others that
have been allocated to two groups. Parentheses indicate whether, in the author’s view, this represented a
Demand Control (DC) or Effort-Reward (ER) variable. The third group is outcome variables, including
measures of three different types of morbidity, drug use, exercise, obesity, and smoking and drinking. Table I
provides a listing of the variables. Note especially that we have considered SES as multidimensional. SES
represents an important human capital variable, education, but it also is an important job characteristic, first
because it represents remuneration from an employer, and second it represents a reward bestowed by society in
terms of prestige.
Human Capital:
1. Highest grade achieved as of 2002
2. Pearlin Mastery – Personality scale administered in 1992
3. Rosenberg Self-Esteem – Personality scale administered in 1980
4. Armed Forces Qualification Test percentile (measured in 1979 at start of study; well correlated with
measures of IQ, and also overall high school grade point average)
5. Number of biological children ever born
6. Reads the labels of prepared food to determine contents
7. Married
Job Characteristics:
1. Hourly rate of pay (natural logarithm of) in 2000 (ER)
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
4
6th Global Conference on Business & Economics
2.
3.
4
5
6
7
8
9
10.
11.
12.
13.
14.
ISBN : 0-9742114-6-X
Prestige (Based on 1980 Occupations and the concordance with the Nakao-Treas prestige ratings
(Hauser & Warren, 1997). (ER)
Exercise (sum of self rated amounts of light and heavy physical exercise) (ER)
Insurance (created variable using summed scalar measures of medical, dental, life insurance, pension,
and job training opportunities) (ER)
Number of hours worked per week (ER)
Number of paid vacation days per year (ER)
Performance pay - interval scale representing amount of worker wages attributable to performance pay
(ER)
Total promotions (ER)
Job autonomy (variable created by self-reports of being in hierarchical relation with boss and
subordinates) (DC)
Flexible work schedule (DC)
Total responsibility (DC)
Total training investment (DC)/
Total training time (DC)/
Informal training with supervisors and coworkers – interval scale with increasing level of weeks of
training with supervisors and coworkers (DC)/
Outcomes:
1. Body Mass Index (BMI), computed as weight in kilograms squared by height in meters
2. Depression (using the CESD scale)
3. Smoking (at least 100 cigarettes in lifetime)
4. Job satisfaction (global job satisfaction in 2000)
5. Number of alcoholic drinks per month
6. Net wealth (created variable by summing all assets including home market value and deducting debt on
home)
7. Drugs 1 (created variable via factor analysis of drug use inventory – reflects the unprescribed use of
tranquilizers, sedatives, stimulants, and painkillers
8. Drugs 2 (created variable via factor analysis of drug use inventory – reflects the unprescribed use of
heroin and syringes
9. Drugs 3 (created variable via factor analysis of drug use inventory – reflects the unprescribed use of
marijuana, cocaine, and crack cocaine.
10. Short Form-12 Physical Health Component (PC) – A linear compound of 12 self-reported health
variables reflecting the overall physical health of the respondent (Ware et al. 1996)
11. Short Form-12 Mental Health Component (MC) – A linear compound of 12 self-reported health
variables reflecting overall mental health of the respondent (Ware et al. 1996).
The two Short Form health measures are in widespread use among medical researchers and using them
allows comparison between this and other studies.
Means and standard deviations and other descriptive statistics are provided in Table I. There were a total of
1297 complete cases. The intercorrelation matrix of these thirty two variables is this matrix that is the input to
our analysis program, LISREL 8.53 (Joreskog and Sorbom, 1996). We have chosen not to reprint it here, but it
is available upon request, as is the entire SPSS file from which it was generated. The data were run only for
those records in which data were complete on all variables.
Results
We began our analysis by testing to make sure that our data would display the positive relationship
between SES and health described above. This seemed to be an important validity criterion our results would
need to satisfy. We computed deciles for the distributions of log wage, education attained, and status scores. To
get a composite SES score we simply added the deciles together for these distributions and assigned each case a
value. We also created 21 occupation categories based on the Census 3-digit 1980 codes for each of the cases in
the 2000 round of the NLSY. Finally, we plotted the mean Short Form components against the SES sc ores for
the occupations. The results are shown in Figures 1-2. The resulting best fit lines show a significant positive
slope, especially pronounced for PC. Consequently, our data and findings corroborate and are consistent with
those from the well-known Whitehall II study (Marmot, et al., 1997). There is a positive relationship between
occupational SES and health in our data.
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
5
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
The best fit LISREL model for the intercorrelation matrix has a chi-square of 367.32 with 355 degrees
of freedom. This yields an overall p-value of 0.351. The AGFI for the model is 0.974 and the RMSEA is
0.00519, indicating a satisfactory fit for the model. Numbers adjacent to causal arrows in the figures 3 -7 are
the t-statistics for each estimate. With few exceptions we have only shown the parameters significant at less
than 0.05 confidence level (t > +/-1.97). Double asterisks indicate p <= 0.01, and single asterisks are for pvalues <= 0.05.
The structural model for the best fitting solution to the intercorrelation matrix is presented in five parts,
in the interest of clarity. As a glance at figures 3 – 7 will readily show, the complexity of the human capital-job
characteristics-health outcomes linkages is extraordinary. We begin by developing a morbidity model for the
data and then show the relationships between human capital variables and the nature of the jobs our respondents
held. Then we show how human capital is related to morbidity by direct linkages. Finally, we present the
relationships we found between effort-reward job characteristics and health, and demand-control variables and
health. We conclude with a brief discussion of our findings and some implications for business people.
1.
A Morbidity Model – see figure 3
We begin by examining how obesity, depression, and drug use (including smoking and alcohol
consumption) are related to physical and mental health, and also how they are related to each other. The main
drivers of physical health are depression and obesity. Obesity is strongly related to physical health and
depression is even more strongly related to physical health. There is a complicated reciprocal relationship
between depression and the mental health component, but the net effect of positive mental health is a reduction
in depression. Net wealth reduces obesity and job satisfaction increases reported mental health levels. Smoking
has a reciprocal relationship with mental health. Smoking improves mental health, but improved mental health
tends to reduce smoking; the net effect, however, is that smoking improves mental health. Smoking also
reduces obesity. The use of Drugs 1 (tranquilizers, stimulants) reduces reported mental health. Marijuana and
cocaine are also positively associated with the use of tranquilizers and stimulants. The main actors in this
complex of behaviors and health outcomes are clearly depression and obesity.
2.
Human Capital variables and Job Characteristics – see figure 4
Having identified the main proximal factors in health outcomes, obesity and depression, we then
sought human capital variables that were related to job characteristics. AFQT and education proved to be by far
the most important, with smaller contributions due to children, self-esteem, and mastery. Education and AFQT
aptitude were strongly related to wages and also to job prestige, two effort-reward variables. Self-esteem and
mastery, personality factors, also increased wages. Children were related to increased levels of exercise, a plus
as we will see, for health outcomes. Marital status had no significant impacts. In summary, education and
AFQT aptitude, the latter strongly correlated with conventional IQ measures such as Wechsler-Bellevue,
modulate wages and job prestige. This is not surprising
3.
Human Capital variables and Health – see figure 5.
Before checking for job characteristic effects on health and health behaviors, we controlled for the
direct effects of human capital on health. As shown in Figure 5, Education is associated with lower levels of
smoking and greater wealth. AFQT is even more strongly associated with net wealth, but it is also related
inversely to obesity and depression. AFQT has offsetting health liabilities. It is associated with elevated levels
of alcohol consumption and with the use of narcotics. Mastery, a personality variable measured about a decade
prior to the health data, also is positively correlated with reported good mental health. We found no significant
relationship between marital status and any of the health measures.
4.
Health Effects of effort - reward variables. See figure 6.
Having controlled for human capital, we turned to effort-reward variables. A number of relationships
between these and health outcomes were observed. Wages were related positively to net wealth, a good
outcome for health since wealth reduces obesity. Higher wages directly impact body mass and reduce it, also a
good outcome. Unfortunately, there is a tendency for higher wages to increase alcohol consumption. Perhaps
this is not surprising since higher wages allow a budget for drinks. Exercise turned out to have a number of
benefits for workers, especially by decreasing obesity. More exercise reduced drinking and smoking. More
exercise was associated with elevated levels of physical and mental health. Occupation prestige was associated
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
6
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
with a lower likelihood of consuming narcotics,. Evidently, job prestige does not favor one’s psychological
health. Insurance helped reduce depression.
Working more hours per week leads to greater BMI levels. Perhaps nibbling on the job is more
pervasive than we might think, and if the jobs are sedentary the extra calories are not reduced by exercise.
Finally, performance pay contributes positively to health, because it impacts net wealth and this reduces obesity.
5.
Health effects of demand– control variables. See figure 7.
The only impact flexible job hours has on any dependent health variable is job satisfaction, where, not
surprisingly, workers in jobs with flexible hours are apparently more satisfied. Inspecting Figure 7 shows that
job autonomy, which is working in a non-hierarchical system, has two effects. First, jobs that are more
autonomous seem to be held by wealthier workers. From a health perspective, this is good because net wealth
leads to lower obesity. However, jobs that are more autonomous are associated with greater use of tobacco. So,
job autonomy has conflicting health benefits.
Discussion
The main concern of this analysis has been to present a structural model of the relationship between
certain social facts about a person, especially their SES and certain conditions of their employment, and there
propensity to become ill. Certain major points emerge. First, SES itself (as indexed by all three of its
components) is positively related to overall physical and mental health in our employee sample. Wages in
particular were directly related to lower obesity and greater net wealth, an important mediator in reducing
obesity. Education also led to more net wealth, and thus lower obesity. Education also reduced smoking levels,
and in our data, lower levels of smoking are associated with higher levels of mental health. The big effect of
education, of course, is that it leads to jobs that have more pay, and therefore increase the chances of improving
net wealth. The prestige element of SES had no relationship with health, although the data show that higher
prestige levels lead to lower use of narcotics. Thus, the major impact of SES is to reduce obesity. What is
interesting is that education does not reduce obesity directly, but through its impact on net wealth; better
educated people have more net wealth and because of this they are thinner.
Controlling for human capital variables allowed us to look at the separate effects of job characteristics.
Here we found that the main actors were mainly effort-reward variables, not demand-control characteristics. We
discussed wages and prestige above. Other effort-reward variables that proved to be significant to health
outcomes were exercise, the number of hours worked, performance pay, number of promotions and insurance
benefits. Of these, the most important was exercise. Exercise reduced obesity, reduced drinking, lowered the
use of marijuana and cocaine, and led to higher reported physical and mental health. Performance pay, an
increasingly popular form of compensation, was significantly related to net wealth, and thus contributed to a
reduction in obesity.
There seemed to be only weak or no effects that could be attributed to demand - control characteristics.
Only two of the demand-control variables played any role. Flexible work hours led to greater reported job
satisfaction, a mediator of reported mental health, but the effect size seems small. More job autonomy
increased net wealth slightly and through this mediator, lowered obesity. However, more job autonomy also
increased smoking. Interestingly, unemployment spells, which we took to indicate a lack of control over job
decisions, was not significantly related to any of our morbidity and health behaviors.
The main precursors to disease in this study are depression and obesity, not SES or other job
characteristics. While there are significant linkages between SES and job characteristics and obesity and
depression, it is clear that the amount of variance attributable to these determinants leaves ample room for other
factors to drive body mass and depression. In summary, job characteristics do matter to health, but one must
look elsewhere to find the major causal agents. Employers can take some steps to improve employee health,
such as creating jobs that are more autonomous and less hierarchical, offering insurance, and providing more
physical exercise as part of the job, and keeping a lid on the number of hours employees work. The data also
show that more vacation and more worker discretion over when they work have little or no value in health
outcomes.
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
7
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
References
Adler, Nancy E., Epel, Elissa S., Castellazzo, Grace, & Ickovics, Jeannette R. (2000). Relationship of Subjective
and Objective Social Status with Psychological and Physiological Functioning: Preliminary Data in
Healthy White Women. Health Psychology, 19, 586-592.
Adler, Nancy E., Boyce, Thomas, Chesney, Margaret A., Cohen, Sheldon, Folkman, Susan, Kahn, Robert L., &
Syme, S. Leonard. (1994). Socioeconomic Status and Health. American Psychologist, Vol. 49, 15-24.
Boehm, Christopher. (1999). Hierarchy in the Forest. Cambridge: Harvard University Press.
Bose, C. D. & Rossi, P.H. (1983). Prestige standings of occupations as affected by gender. American
Sociological Review, 48, 316 -330.
Bosma, Hans, Peter, Richard, Siegrist, Johannes, and Marmot, Michael. (1998). Two Alternative Job Stress
Models and the Risk of Coronary Heart Disease. American Journal of Public Health, Vol. 88, 68-74.
Brunner, E. J., Marmot, M.G., Nanchahal, K., Shipley, M.J., Stansfeld, S. A., Junejn, M., & Alberti, K.G.M.M.
(1997). Social inequality in coronary risk: central obesity and the metabolic syndrome. Evidence from
the Whitehall II study. Diabetologia, Vol. 40, 1341-1349.
Cassidy, John. (2006). Relatively Deprived. The New Yorker, 82, 42-47.
Davey Smith, George, Shipley, Martin J., & Rose, Geoffrey. (1990). Magnitude and causes of socioeconomic
differentials in mortality: further evidence from the Whitehall Study. Journal of Epidemiology and
Community Health, 44, 265-270.
Feinstein, Jonathan S. (1993). The Relationship between Socioeconomic Status and Health: A Review of the
Literature. The Milbank Quarterly, Vol. 71, 279-322.
Gallo, Linda & Matthews, Karen A. (2003). Understanding the Association between Socioeconomic Status and
Physical Health. Psychological Bulletin, Vol. 129, 10-51.
Hauser, Robert M., & Warren, John Robert. (1997). Socioeconomic Indexes for Occupations: A Review,
Update, and Critique. Sociological Methodology, 27, 177-298.
Kawachi, Ichiro. (2000). Income Inequality and Health. In Berkman, Lisa F. & Kawachi, Ichiro (Eds.) Social
Epidemiology. Oxford: Oxford University Press. 76-94.
Kennedy, Bruce P., Kawachi, Ichiro, Glass, Roberta, & Prothrow-Stith, Deborah. (1998). Income distribution,
socioeconomic status, and self rated health in the United States: multilevel analysis. British Medical
Journal, Vol. 317, 917-921.
Lynch, John W., Davey Smith, George, Kaplan, George A., & House, James S. (2000). Income inequality and
mortality: importance to health of individual income, psychosocial environment, or material conditions.
British Medical Journal, Vol. 320, 1200-1204.
Lynch, John, Krause, Niklas, Kaplan George A., Tuomilehto, Jaakko, & Salonen, Jukka T. (1997). Workplace
Conditions, Socioeconomic Status, and the Risk of Mortality and Acute Myocardial Infarction: The
Kuopio Ischemic Heart Disease Risk Factor Study. American Journal of Public Health, Vol. 87, 617622.
Lynch, J. W., Kaplan, G. A., & Salonen, J. T. (1997). Why do Poor People Behave Poorly? Variation in Adult
Health Behaviours and Psychosocial Characteristics by Stages of the Socioeconomic Lifecourse. Social
Science & Medicine, Vol. 44, 809-819.
Lynch John W., & Kaplan, George A. (1997). Understanding how inequality in the distribution of income
affects health. Journal of Health Psychology, Vol. 2, 297-314.
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
8
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
Marmot, Michael. 2004. The Status Syndrome. New York: Henry Holt and Company.
Marmot, Michael and Wilkinson, Richard G. 1999 (Eds.). Social Determinants of Health. Oxford: Oxford
University Press.
Marmot, Michael, Ryff, Carol D., Bumpass, Larry L., Shipley, Martin, and Marks, Nadine F. (1997). Social
Inequalities in Health: Next Questions and Converging Evidence. Social Science and Medicine, 44,
901-910.
Melamed, Samuel, Shirom, Arie, Toker, Sharon, Berliner, Shlomo, & Shapira, Itzhak. (2006). Burnout and Risk
of Cardiovascular Disease: Evidence, Possible Causal Paths, and Promising Research Directions.
Psychological Bulletin, Vol. 132, 327-353.
Ostrove, Joan M., Adler, Nancy E., Kuppermann, Miriam, & Washington, A. Eugene. (2000). Objective and
Subjective Assessments of Socioeconomic Status and Their Relationship to Self-Rated Health in an
Ethnically Diverse Sample of Pregnant Women. Health Psychology, 19, 613-618.
Ross, Nancy A., Wolfson, Michael C., Dunn James R., Berthelot, Jean-Marie, Kaplan, George A., & Lynch,
John W. (2000). Relation between income inequality and mortality in Canada and in the United States:
cross sectional assessment using census data and vital statistics. British Medical Journal, Vol. 320,
898-902.
Sapolsky, Robert M. 2004. Why Zebras Don’t Get Ulcers. New York: Henry Holt and Company.
Sapolsky, Robert M. (1990). Adrenocortical Function, Social Rank, and Personality among Wild Baboons.
Biological Psychiatry, Vol. 28, 862-878.
Siegrist, J. (1996). Adverse health effects of high-effort/low-reward conditions. Journal of Occupational Health
Psychology, Vol. 1, 27-41.
Singh-Manoux, Archana, & Marmot, Michael. (2005. Role of socialization in explaining social inequalities in
health. Social Science & Medicine, Vol. 60, 2129-2133.
Singh-Manoux, Archana, Adler, Nancy E., & Marmot, Michael. (2003). Subjective social status: its
determinants and its association with measures of ill-health in the Whitehall II study. Social Science &
Medicine, 56, 1321-1333.
Theorell, T., & Karasek, R. A. (1996). Current issues relating to psychosocial job strain and cardiovascular
disease research. Journal of Occupational and Health Psychology, Vol. 1, 9-26.
Theorell, Tores, Tsutsumi, Akizumi, Hallquist, Johan, Reuterwall, Christina, Hogstedt, Christer, Fredlund,
Peeter, Emlund, Nils, Johnson, Jeffey V., & the SHEEP Study Group. (1998). Decision Latitude, Job
Strain, and Myocardial Infarction: A Study of Working Men in Stockholm. American Journal of Public
Health, Vol. 88, 382-388.
Ware JE, Kosinski M, and Keller SD. (1996). A 12-Item Short-Form Health Survey: Construction of scales and
preliminary tests of reliability and validity. Medical Care, 34, 220-233.
Wilkinson, R. G. (1999). Putting the picture together: prosperity, redistribution, health, and welfare. In Marmot,
Michael & Wilkinson, Richard G. (Eds.) Social Determinants of Health. Oxford: Oxford University
Press. 256-274.
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
9
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
Table 1
Means and Standard Deviations
HIGHEST GRADE COMPLTD 2002
PEARLIN
ESTEEM
PROFILES AFQT PRCTILE 89 (REV) 81
NUMBER OF CHILDREN EVER BORN 2002
READ NUTRITIONAL INFO ON FOOD 2002
MARRIED
LNWAGE00
PRESTIGE
EXERCISE
INSURANC
HOURS PER WEEK WORKED L1 2000
# VACATION DAYS R GETS PER YR? L1 2000
PERFPAY
TOTPROMO
JOBAUT
FLEXIBLE WORK SCHEDULE FROM EMP L1 2000
TOTRESP
TRTOTINV
TRTOT
TOTUNEMP
BMI
CESD2002
SMOKD AT LEAST 100 CIGRTS IN LIFE? 1998
LIKE JOB MUCH,FRLY WELL,DSLIKE L1 2000
NUMDRINK
WEALTH
DRUGS1
DRUGS2
DRUGS3
SF-12 SCORE, PCS 2002
SF-12 SCORE, MCS 2002
Figure 1
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
10
Mean
Std. Deviation
13.650
2.492
1.796
0.427
1.594
0.394
48.520
29.371
1.910
1.390
2.810
1.571
0.650
0.477
7.344
0.574
45.353
13.491
2.551
1.124
0.766
0.318
42.270
10.310
13.800
15.436
0.045
0.099
0.553
0.762
0.504
0.627
0.540
0.499
1.139
1.056
12.278
22.230
0.739
0.904
8.211
23.626
28.367
5.596
0.391
0.506
0.470
0.500
1.630
0.694
4.463
6.982
125505.173 273863.772
0.111
0.230
0.012
0.095
0.300
0.281
5305.030
673.041
5407.840
692.327
N
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
1297
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
PCS
SF-12 PCS vs Socioeconomic Status
55
54.5
54
53.5
53
52.5
52
51.5
51
50.5
y = 0.0348x + 50.515
R2 = 0.4789
SF-12 PCS
Linear (SF-12 PCS)
0
20
40
60
80
100
SES
Figure 2
SF-12 MCS vs Socioeconomic Status
56
y = 0.0156x + 52.35
R2 = 0.0891
MCS
55
54
SF-12 MCS
53
Linear (SF-12 MCS)
52
51
50
0
20
40
60
80
SES
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
11
100
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
Figure 3
A Morbidity Model for NLSY 79 Data
1.09*
drugs 1
drugs2
-.08**
BMI
.55**
-.08**
-.07*
drugs3
.07*
-.06*
.13**
.35**
CESD
Drink
.15**
-.78**
.22**
Smoke
-.28**
-.19**
wealth
pcs
-.38**
.40*
-.07**
-.22**
jobsat
mcs
-.09**
.48**
-.20**
Figure 4
Relationships between Exogenous and Certain Job Characteristic
Variables
Logwage
numkids
prestige
exercise
insurnce
.34**
.28**
.08**
education
.09*
-.11**
-.09**
self-est
.06*
jobaut
flexhours
-.09**
mastery
totresp
-.08*
.08*
readlabl
trtotinv
.32**
.05*
.20**
.08**
AFQT
Married
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
12
totrain
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
Figure 5
Exogenous Variables & Health Outcomes
Numkids
MCS
wealth
jobsat
-.26
.09**
.12**
.09**
education
drugs3
.10**
.12**
Self-est
Drink
.12**
Mastery
BMI
-.09**
Readlbl
-.20**
-.13**
AFQT
.07**
CESD
Smoke
Married
Figure 6
Relationships between Effort/Reward Job Characteristics
and Health Outcomes
.21**
logwage
wealth
BMI
-,06*
drugs2
.15**
hourswork
.10**
-.10**
drugs3
-.08**
prestige
.12**
-.09**
exercise
.06*
Drink
Smoke
-.07**
vacation
CESD
-.15**
.09**
.09**
pcs
-.12*
perf pay
mcs
totpromo
insurnce
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
13
jobsat
6th Global Conference on Business & Economics
ISBN : 0-9742114-6-X
Figure 7
Relationships between Demand/Control Job Characteristics
and Health Outcomes
.10**
jobaut
wealth
BMI
flexhours
drugs1
drugs2
drugs3
totresp
Drink
.08**
totrninv
Smoke
CESD
totrain
-.12**
pcs
mcs
unemploy
jobsat
OCTOBER 15-17, 2006
GUTMAN CONFERENCE CENTER, USA
14